Summary: New research shows that poor sleep reduces the accuracy of facial identification and may increase the risk of confident but incorrect identifications.
Source: University of Glasgow.
Identifying unfamiliar people by comparing face images is a common and vital task—for example matching CCTV images to a mugshot or checking a passport photograph against a traveller at a border.
Researchers from the University of New South Wales, in collaboration with the University of Glasgow, report in Royal Society Open Science that inadequate sleep impairs accuracy on face-identification tasks. The study examined face matching under conditions that remove memory demands and better reflect routine security and forensic checks.
The study used a standard face-matching test in which participants decided whether two images shown side-by-side depicted the same person or two different people. Two key features distinguished this task from everyday face recognition: the faces were unfamiliar to participants, and the images were presented simultaneously, so the task did not rely on memory.
Prior research has established that restricted sleep impairs memory for faces. What remained unclear was whether poor sleep would also harm perceptual face identification when memory is not involved. This study addressed that gap and found that sleep quality affects the perceptual processes involved in matching faces.
Dr Louise Beattie of the University of Glasgow’s School of Psychology explained that participants who reported disrupted sleep over the three days before testing performed worse on the Glasgow Face-Matching Task than those with normal sleep patterns. In a separate experiment, participants with symptoms consistent with insomnia also showed reduced accuracy on the same face-matching measure.
The researchers emphasize two practical implications. First, sleep disruption is widespread in the general population and is particularly common among night-shift workers. Second, many security, border control, and policing roles require repeated photo-ID checks under shift-work schedules. The study shows that poor sleep can reduce accuracy in critical “passport-type” tasks, potentially increasing the likelihood of mistaken identity.
Importantly, the study found that reduced accuracy was not accompanied by reduced confidence. David White from the University of New South Wales noted that while poor sleep correlated with lower face-matching accuracy, participants experiencing poor sleep reported confidence in their answers at similar levels to well-rested participants. That combination—higher error rates with unchanged confidence—carries clear risks for occupational settings where confident misidentifications can have serious consequences.
The authors suggest that sleep-related declines in face memory reported previously may reflect underlying perceptual encoding difficulties: poor sleep appears to impair the visual processing needed to encode and compare facial identity, rather than only degrading later memory retrieval. The findings also point to possible metacognitive impairments: people who are sleep-deprived may be less aware of their reduced performance.
Funding: The research was supported by the Australian Passport Office, the Australian Research Council, and the Economic and Social Research Council.
Research team and publication: The study, titled “Perceptual impairment in face identification with poor sleep,” was authored by Louise Beattie, Darragh Walsh, Jessica McLaren, Stephany M. Biello, and David White and published in Royal Society Open Science (published online October 5, 2016).
Key findings:
- Poor sleep over several nights is associated with lower accuracy on face-matching tasks that do not rely on memory.
- Participants reporting insomnia symptoms performed worse on the Glasgow Face-Matching Task.
- Reduced accuracy was not matched by reduced confidence; poor sleepers remained as confident in their (often incorrect) decisions as well-rested participants.
- Results highlight risks for security, border control, and policing professions that depend on accurate photo-ID checks, especially under shift-work conditions.
These results build on the broader literature linking sleep disruption and shift work to adverse cognitive, emotional, and health outcomes. They underline the importance of considering sleep quality in occupations requiring precise perceptual judgements and suggest that interventions to improve sleep or adjust work schedules could reduce identification errors in operational settings.
The research tested whether poor sleep affects performance on a face-matching test that does not require memory: the Glasgow Face-Matching Task (GFMT). In two experiments, participants with self-reported sleep disturbance or short sleep over consecutive nights showed reduced GFMT accuracy compared with well-rested controls. Across both experiments, reduced accuracy in poorer sleepers was not accompanied by lower confidence, raising important occupational and safety concerns. The findings indicate that sleep-related face recognition problems arise from perceptual encoding difficulties and suggest metacognitive impairment following poor sleep.
For media and educational use, this summary is based on the University of Glasgow press release and the published open-access article in Royal Society Open Science.